syndicated loan spreads and the composition of the syndicate · · 2012-12-14syndicated loan...
TRANSCRIPT
Syndicated Loan Spreads and the Composition of the Syndicate
Jongha Lim
University of Missouri
Bernadette A. Minton
Ohio State University
Michael S. Weisbach
Ohio State University, NBER, and SIFR
August 28, 2012
Abstract
The past decade has seen significant changes in the structure of the corporate lending market, with non-bank
institutional investors playing larger roles than they historically have played. These non-bank institutional lenders
typically have higher required rates of return than banks, but invest in the same loan facilities. We hypothesize that
non-bank institutional lenders invest in loan facilities that would not otherwise be filled by banks, so that the
arranger has to offer a higher spread to attract the non-bank institution. In a sample of 20,031 leveraged loan
facilities originated between 1997 and 2007, we find that, loan facilities including a non-bank institution in their
syndicates have higher spreads than otherwise identical bank-only facilities. Contrary to risk-based explanations of
this finding, non-bank facilities are priced with premiums relative to bank-only facilities of the same loan package.
These premiums for non-bank facilities are substantially larger when a hedge or private equity fund is one of the
syndicate members. Consistent with the notion that firms are willing to pay spread premiums when loan facilities are
particularly important to the firm, we find that firms spend the capital raised by loan facilities priced at a premium
faster than other loan facilities, especially when the premium is associated with a non-bank institutional investor.
Contact information: Jongha Lim, Department of Finance, University of Missouri, email: [email protected];
Bernadette A. Minton, Department of Finance, Fisher College of Business, Ohio State University, Columbus, OH
43210: email: [email protected]; Michael S. Weisbach, Department of Finance, Fisher College of Business,
Ohio State University, Columbus, OH 43210, email: [email protected]. We would like to thank Zhenyi
Huang and Jongsik Park for excellent research assistance, and participants in a seminar at University of Missouri,
Ohio State University, and Penn State University, as well as Zahi Ben-David, Isil Erel, John Howe, Wei Jiang, Kai
Li, Robert Prilmeier, Shastri Sandy, Berk Sensoy, Pei Shao, René Stulz, and Jun Yang for helpful suggestions.
1
1. Introduction
Various types of institutional investors participate in syndicated loans. These investors have
substantially different costs of providing debt capital: Banks must receive the risk free rate plus a
premium for the default risk. In contrast, hedge fund managers have relatively high required returns on
top of the considerable fees they charge. Consequently, to justify it making an investment, a hedge fund’s
pre-fee expected returns must be substantially higher than those for a bank. Given these different expected
returns, it is somewhat puzzling that both hedge funds and banks (as well as other institutions) all invest
in the same syndicated loan facilities.
One possible explanation for the observation that investors with differing expected returns invest
in the same syndicated loan facilities is that facilities differ on dimensions other than risk, and that these
differences are associated with both spread differences and also the identity of investors who provide the
financing. Some loan facilities are made when the supply of capital is high, so that the loan facility can be
filled by banks at a relatively low spread. Others are made at times when it is difficult to acquire the
necessary capital from banks, so that the loan arrangers have to raise the spreads to attract other non-bank
institutional investors such as hedge funds. Alternatively, if the loan facility is not crucial to the firm's
health and it cannot be filled at low cost by banks, the firm could choose not to borrow at all.
Consequently, when non-bank financial institutions take positions in loan facilities, there should be a
higher spread than in loan facilities in which they do not take positions. In addition, we expect that
borrowing firms should spend the money they raise in non-bank facilities relatively quickly.
To evaluate the way in which different kinds of non-bank institutional investors are involved in
the syndicated lending process, we consider a sample of 20,031 facilities of “leveraged” loans from the
DealScan database, each of which was originated between 1997 and 2007.1 We focus on the leveraged
1 The technical definition of leveraged loans varies by organization. For example, DealScan defines as leveraged
any loan with a credit rating of BB+ or lower and any unrated loan. Bloomberg defines leveraged loans as those
with spreads over LIBOR of 250 basis points (bp) or more. Standard & Poor’s deems loans with spreads over
LIBOR of 125 bp or more as leveraged loans. Thompson Financial denotes as leveraged loans, all those with an
initial spread of 150 bp or more before June 30, 2002, or 175 bp or more after July 2, 2002. We follow DealScan’s
2
loan segment of the market because non-bank institutional investors’ participation in the corporate
lending market has been concentrated in this lower quality, non-investment grade segment of the market,
and also because restricting the sample to leveraged loans allows us the sample to be relatively
homogenous.2 Of the 20,031 leveraged loan facilities, 13,752 are associated with a syndicate containing
only commercial or investment banks (bank-only facilities), while the remaining 6,279 have syndicates
containing at least one non-bank institutional investor (non-bank facilities). These institutional investors
are most often finance companies (contributing to the syndicates of 4,603 loan facilities), private equity or
hedge funds (2,754 loan facilities) and mutual funds (1,010 loan facilities).
We estimate the difference in spreads between loan facilities as a function of the identities of the
investors in a particular facility. In doing so, we control for other factors that affect the loan facility’s
spread, such as the firm’s risk measured by either firm-level accounting variables, or the rating of the
issuer, as well as the loan facility’s type (Term Loan A, Term Loan B, or Revolver) and other facility-
specific characteristics. Our estimates suggest that the presence of a non-bank institutional syndicate
member is associated with a significantly higher spread than an otherwise similar bank-only loan facility.
When we control for risk using firm-level accounting variables, our estimates imply a spread premium of
approximately 56 basis points. If we instead group loans by rating category, the estimated spread
premiums are smaller, around 24 basis points, but are still statistically significant and large enough to be
economically important.
In computing these estimates of the non-bank premiums, we control for publicly observable
variables that could affect spreads. However, it is possible that non-bank premiums could reflect
unobservable differences between firms that are correlated with both the likelihood of there being a non-
classification of leveraged loans in this paper. By “non-bank” we mean an institutional investor that is neither
commercial bank nor investment bank. 2 The proportion of leveraged loans among loans classified as “institutional” loans by DealScan is about 90% during
the sample period. Similarly, Nandy and Shao (2010) find that 86.1% of “institutional” loans are leveraged loans
with the proportion increasing over the years during the period from 1995 to 2006. The definition of “institutional”
facilities in this paper is different from the one used by DealScan or Nandy and Shao (2010). We focus on the actual
participation as opposed to the label put on the facility and consider a loan facility to be ‘institutional’ if at least one
non-bank (neither commercial bank nor investment bank) institutional investor is involved in the lending syndicate.
3
bank institutional syndicate member and the spreads on the loan facilities in which they invest. For
example, suppose that at times when the firm is having financial problems that prevent it from receiving a
loan facility from other lenders, it is more likely to have a non-bank institution in the loan facility’s
syndicate. In this case, it would be possible that the borrower’s true risk would not be reflected in
observable variables, so that the positive estimated premiums could reflect compensation for risk that is
unobservable to an outsider.
To evaluate the possibility that the premiums to non-bank institutional investors reflect
incremental risk differences between non-bank loans facilities and bank-only loan facilities, we estimate
the effect of non-bank syndicate members on the pricing of different facilities within the same loan.
Different facilities within the same loan package typically have the same seniority and hence have the
same default risk. Yet, facilities usually have different maturities, sizes, and syndicate structures, so we
control econometrically for differences in facility-specific attributes when estimating within-loan
differences. Using this approach, the existence of a non-bank syndicate member effect on the relative
spreads on different facilities of the same loan cannot reflect a correlation between non-bank institutions’
existence and a factor related to firm-level risk.
The within-loan estimates indicate that when a non-bank institution participates in a Term Loan B
facility’s syndicate, the facility has a larger spread premium relative to Term Loan A facilities or
revolvers of the same loan relative to bank-only Term Loan B facilities, although only the premium
difference for revolvers is statistically significantly different from zero. We also consider the cases in
which the non-bank institution invests in a particular type of facility and there also is another facility of
the same type in the same loan. In each of these cases, the facility with the non-bank institutional investor
trades at a statistically significantly higher spread. These findings confirm that facilities in which non-
bank institutional investors participate have higher spreads than otherwise similar bank-only facilities,
even holding borrower characteristics constant.
We also examine whether the type of non-bank syndicate member is related to the spread
premium. We estimate this spread premium when we control for risk econometrically using firm-specific
4
financial data, and also when we compare across different facilities in the same loan. Consistent with the
notion that different types of institutional investors have different required rates of return, we find that
when hedge or private equity funds participate in a facility’s syndicate, the spread premium is
substantially higher, about 29 basis points than when other types of non-bank institutional investors
participate in the facility’s syndicate.
We also examine whether these spread premiums vary when the non-bank syndicate members
also have equity positions at the time of the loan facility origination. When a hedge fund has an equity
stake in the borrowing firm greater than 0.1 percent, the spread premium approximately doubles, to about
58 basis points. Finally the spread premiums vary positively with the fraction of the loan that is purchased
by the non-bank institutional investor. These findings are consistent with the view that arrangers rely on
non-bank institutional investors, especially hedge and private equity funds, as lenders of last resort, when
it is difficult to raise capital for the loan facility through banks.
Finally, we consider the idea that if non-bank institutional lenders are paid premiums to invest
in the particular facilities in which they participate, then these facilities should be for loan facilities in
which capital is particularly important for the borrowing firms. A testable implication of this idea is that
the borrowing firms should save a smaller fraction of the capital they raise as cash than when a non-bank
institution does not participate in the syndicate. To evaluate this hypothesis, we estimate equations similar
to those in Kim and Weisbach (2008) that predict the fraction of an incremental dollar raised that goes to
alternative uses. Our estimates indicate that when there is a positive spread premium, the estimated
fraction of capital raised that the borrowing firm saves as cash declines with the abnormal spread on the
loan facility. This finding is consistent with the notion that borrowers are willing to pay a premium on
their loan facilities in situations in which raising capital quickly is particularly important to the firm.
Our findings parallel those of Brophy, Ouimet, and Sialm (2009), who find that hedge funds’
equity investments are typically to firms that otherwise would have trouble raising capital. When making
equity investments, hedge funds typically negotiate discounts relative to the public stock price paid by
other investors, and earn abnormal returns because their purchases are at a discount. Thus, hedge funds
5
abnormal returns on private placements of equity can be thought of as the return to providing liquidity.
Our findings can be viewed similarly: we find that hedge and private equity funds contribute to loan
facilities in firms with spreads that are relatively high. Since spreads are determined through an auction
process, high spreads that cannot be explained by risk and other firm and loan facility attributes mean that
the facility would have relatively few investors or would have difficulties in fully subscribing absent the
hedge or private equity fund. Therefore, we view the spread premiums as compensation that non-bank
institutional investors receive in exchange for providing liquidity to their firms in the facility that is in less
demand from other investors.
Nandy and Shao (2010) compare spreads on “institutional” and “bank” facilities, and document
that the Term Loan B facilities or what they label institutional facilities, have higher spreads than
facilities they label bank facilities, Revolvers or Term Loan A facilities. We extend their work in a
number of ways; in particular, we focus on the actual participation by types of bank and non-bank
institutional syndicate members as opposed to the label put on the facility, and the way that non-bank
institutions participation in the syndicate affects facilities’ spreads and the way in which the capital is
used.
The paper also is related to the literature on potential conflicts of interest that arise when
institutional investors engage in syndicated lending. Ivashina and Sun (2011b) and Massoud, Nandy,
Saunders, and Song (2011) focus on the trading of institutions that participate in syndicated lending, and
the associated potential conflicts of interest. Both papers find evidence that institutional investors in the
syndicated loan market exploit their access to private information when trading and earn abnormal returns
when they trade in the firm’s equities.
The remainder of the paper proceeds as follows: Section 2 describes the data sources and
sample. Section 3 estimates the differences in spreads between bank-only loan facilities and comparable
non-bank loan facilities. Section 4 examines factors that affect the magnitude of spread differences
between bank-only and non-bank facilities. Section 5 considers the way in which the fraction of the
capital raised varies with the loan’s pricing and composition of the syndicate, while Section 6 concludes.
6
2. Data sources and sample construction
2.1. Sample of leveraged loan facilities
We obtain our sample of leveraged loans from the Reuters Loan Pricing Corporation’s (LPC)
DealScan database for the 1997-2007 period. We consider a loan to be a “Leveraged loan” if it has a
credit rating of BB+ or lower, or is unrated (see footnote 1). Leveraged loans in our sample are either
stand-alone facilities (41.4%) or made up of term loans facilities packaged together with revolver
facilities. A term loan facility is a loan facility for a specified amount, fixed repayment schedule and
maturity, and is usually fully funded at origination. In contrast, revolvers typically have shorter maturities
than term loan facilities and are drawn down at the option of the borrower. Term Loan facilities are
normally designated by letter, where the Term Loan A facility is usually amortizing, and is typically held
by the lead arranger, and the remaining facilities (Term Loan B, C, D, E, …) are more often “bullet”,
meaning that they have one payoff at maturity, and are usually sold to third parties.3
We focus on leveraged loans because participation of non-bank institutional investors in this
segment of the loan market has increased over time. In addition, according to previous studies (e.g. Nandy
and Shao (2010)) the overwhelming majority of “institutional” loans are leveraged loans with the
proportion increasing over time. Moreover, given that the pricing function of leveraged loans is likely to
be different from that of investment grade loans, restricting the sample to leveraged loans finesses
econometric difficulties that could potentially arise if we were to pool leveraged and investment grade
loans. We begin our sample in 1997 because major developments in the market that fueled institutional
involvement in the corporate loan market occurred in 1995 and 1996.4
3 Appendix C contains statistics on the payoff structure of each type of facility in our sample.
4 The Loan Syndications and Trading Association (LSTA) was founded in 1995 and S&P first started rating bank
loans in 1995. In 1996, LSTA first started providing mark-to-market pricing (for dealers only). In addition, the
secondary market for syndicated loans became well established by mid-1990s: by the early 1990s specialized loan
trading desks were operating in a number of institutions led by Bankers Trust, Alex Brown, Bear Stearns, Citibank,
Continental Bank and Goldman Sachs. By 1997, about 25 institutions had active trading desks and there were two
inter-dealer brokers. These innovations spurred the fast growth of the syndicated loan market, which in turn fueled
institutional participation in the primary lending market. Moreover, there are very few leveraged loans before 1997.
7
To construct the sample, we begin with all leveraged loan facilities listed in DealScan made to
non-financial U.S. public firms and completed between 1997 and 2007, a total of 37,552 loan facilities.
We require that the data on deal value and the date of origination not be missing, and that interest rate is
set at a spread over LIBOR. We additionally restrict the sample to the most common type of facilities,
where the type of instrument is either a line of credit (such as Revolver/Line, 364-Day Facility, Limited
Line) or a term loan.5 We further restrict the sample to the borrowing companies for which we could
match to the Compustat database.6 Finally we exclude loans whose primary purpose is LBO financing.
This screening process results in a sample of 20,031 facilities, associated with 13,122 loans made to 5,627
borrowing firms.
We consider a loan facility to have a non-bank institutional investor if at least one institutional
investor that is not either a commercial or investment bank is involved in the lending syndicate. Non-bank
institutions include hedge funds, private equity funds, mutual funds, pension funds and endowments,
insurance companies, and finance companies.
To identify commercial bank lenders, we start from lenders whose type in DealScan is “US
Bank”, “African Bank”, “Asian-Pacific Bank”, “Foreign Bank”, “Eastern Europe/Russian Bank”,
“Middle Eastern Bank”, “Western European Bank”, or “Thrift/S&L”. We manually exclude the some
observations that are classified as a bank by DealScan but actually are not, such as GMAC Commercial
Finance. Then we manually check lenders whose primary SIC code fall in 6011-6082, 6712, or 6719 and
add them to the list of commercial banks if appropriate. When identifying commercial banks, we also
consider finance companies affiliated with commercial banks (e.g. Foothill Capital) to be commercial
banks. We do take into consideration the changes in the institutional type, so that, for example, JP
Morgan is classified as an investment bank before its merger with the Chase Manhattan Corp in 2000, and
JP Morgan Chase is coded as a commercial bank afterward.
5 This restriction excludes bankers’ acceptance, leases, standby letters of credit, step payment leases, guidance lines,
traded letters of credit, multi-option facilities, and undisclosed loans. 6 We are grateful to Michael Roberts for providing the Dealscan-Compustat link file. In addition to using this link
file, we also manually confirmed the matching between DealScan and Compustat.
8
To identify investment banks, we start from lenders that are classified by DealScan as
“Investment Bank”. By manually checking each lender, we drop ones that are labeled as IB by DealScan
but are better characterized as other types, allowing us, for example, to classify Blackstone Group as a
private equity firm rather than as an investment bank. We also manually check lenders whose primary
SIC code is 6211 to capture additional IB lenders such as RBC Capital Market. Insurance companies are
identified following similar process, focusing on the lenders labeled as “Insurance Company” by
DealScan and the ones having primary SIC code of 6311-6361, 6399, or 6411.
Identifying other types of lenders is more challenging, since there are not SIC codes clearly
indicating finance companies, mutual funds, or hedge funds and private equity funds. Therefore, to
identify finance companies, we rely on DealScan’s classification (“Finance companies”). A lender is
classified as a mutual fund if its type in DealScan is either “Mutual funds” or “Institutional investor-
prime funds”. When a lender’s type in DealScan is ambiguous (e.g. “Institutional Investor – Other”, or
“Other”), we further check Capital IQ to see whether it is a mutual fund. Finally, to identify hedge funds
and private equity funds, we start from the lenders that are labeled as “Institutional investor – Hedge fund”
or “Vulture fund” in DealScan. A lender is further added to the category of HF/PE if its name appears in
the TASS or Preqin databases, or if the descriptions of the lender in Capital IQ tells that it is privately
owned hedge fund sponsor, manages private equity funds, or manages assets for high-net worth
individuals.
Because our sample only includes loan facilities with floating-rate interest payments, we use
the all-in-drawn spread as our measure of loan pricing. The all-in-drawn spread is the sum of the spread
of the facility over LIBOR and any annual fees paid to the lender group. DealScan also provides data on
the facility’s size and maturity, the number of investors participating in the lending syndicate, as well as
information on whether the facility is senior, secured, second-lien, syndicated, and the type of facility
(revolver or term loan). We also consider the firm’s lending relationship with the facility’s syndicate
members by examining whether the firm borrowed from the same lender previously.
9
We match the borrower’s and/or borrower’s parent name to the Compustat firm by a
combination of algorithmic matching and manual checking following Chava and Roberts (2008). Using
this matching procedure, we are able to obtain other firm-level variables from Compustat, CRSP, I/B/E/S,
13F, and SDC Platinum. The total number of leveraged loan facilities that have a full set of data for the
most recent prior fiscal year-end is 12,346, of which 3,460 have participation of an institutional investor
that is neither an investment bank nor a commercial bank.
2.2. Overview of sample
Table I provides statistics on the annual distribution of leveraged loan facilities. This table
emphasizes the increasing trend of non-bank institutional participation in the leveraged loan market. The
value of loan facilities with non-bank syndicate members, as well as the fraction of all leveraged loan
facilities made up by loan facilities with non-bank participation increased substantially over our sample
period, from $57 billion (19% of all leveraged loans) in 1997 to $110 billion (32%) in 2007.
Table II presents summary statistics for the all loan facilities in our sample (20,031 facilities)
(Panel A) and lender participation (Panels B, C, and D). As reported in Panel A, the average facility
amount is $158 million, the average number of investors involved in a lending syndicate is about six, and
the average maturity is about 47 months. Approximately 70% of facilities are secured, 1.4% are second-
lien, and about 87.6% are syndicated loan facilities.7 Almost all loan facilities are senior debt.
Panel B of Table II presents the frequency of bank and non-bank institutions participation, while
Panels C and D report loan share information in a sample of loan facilities for which we have data on loan
shares (5,624 facilities, about 25% of the sample). For this sub-sample, non-bank institutions participate
in about 23% of the loan facilities (1,282 of 5,624). When they participate in the loan facility, non-bank
syndicate members together own 44% of the facility, with finance companies and hedge/private equity
funds each owning about a third of the loan facilities in which they invest. In addition, when non-bank
institutional investors participate in a loan facility, they are the largest investor 46% of the time (Panel D).
7 All results are similar when we exclude sole-lender loans from the sample.
10
The Term Loan A facility often is referred to as the bank facility and Term Loan B facility as the
institutional facility (see for example Nandy and Shao (2010)). However, our data indicate this
description can be somewhat misleading since non-bank institutions do invest in Term Loan A and
revolver facilities as well, and sometimes Term Loan B facilities are held entirely by banks. As Panels B
through D show, contrary to the common terminology, non-bank institutions invest in Term Loan A
facilities and banks invest in Term Loan B facilities. Conditional on investing in Term Loan A facilities
non-bank institutions together own 25% of the facility and when bank invest in Term Loan B facilities,
they together own 86% of the loan facility.
3. Differences between bank-only and non-bank loan facilities
3.1. Univariate differences
Table III summarizes the univariate differences between the 6,279 non-bank and 13,752 bank-
only loan facilities in our sample. Non-bank facilities are less likely to be revolvers than bank-only
facilities loans (49.3% vs. 67.8%), and this difference is statistically significant at the one percent level.
The remainder of the loan facilities in the sample are Term Loans, so non-bank facilities are more likely
to be Term Loan facilties than are bank-only facilities. Facilities designated “Term Loan A” are usually
amortizing, while those “Term Loan B” more often have one final “bullet” payment.8
Within the sample of leveraged loan facilities, the non-bank facilities tend to be more risky than
bank-only facilities. Of the borrowers that do have ratings, non-bank facilities tend to have borrowers
with lower ratings.9 For example, of the non-bank loan facilities with issuer ratings, 56% have a B rating
8 We treat facilities with B or higher designations (e.g. C, D, etc.) as Term Loan B. Moreover, about 49% of the term
loans in our sample has no letter designation but is just called ‘Term Loan’. In all reported tables, we treat these
undesignated term loans as Term Loan B. We do so because the facility attributes, such as the spread and payment
schedule, of the unlabeled Term Loans in our sample appear to be more like the Term Loan B’s than the Term Loan
A’s. Detailed comparisons of attributes across different facility types are provided in Appendix C. In addition, when
a facility is first launched and appears in the ‘Calendar’, which is the weekly record of outstanding loans published
by Reuters Loan Pricing Corporation (LPC), often its type is originally described as “Term Loan”, but ultimately is
classified in DealScan as “Term Loan B”, or vice versa. We have re-estimated all equations reported in the paper
treating unclassified term loans separately and all results are similar to those reported below. 9 We use issuer rating as of the fiscal year-end prior to the loan origination, not ratings for individual loans, because
information on ratings for individual loans is more often missing. Therefore the sample includes 1,100 loan facilities
11
or lower, compared to 36% of the bank-only loan facilities. In addition borrowers of non-bank facilities
have higher leverage, a lower Z-score, are more likely to have negative net worth, and lower ROA than
bank-only facilities.10
3.2. Differences in spreads
The goal of this paper is to understand why we observe investors with different required returns
investing in the same syndicated loan facilities. Within a particular facility, all investors receive the same
return; however, facilities differ cross-sectionally, both in terms of the syndicate composition and the
spreads that they offer investors. To attract investors with a higher required rate of return, facilities must
offer higher spreads. Therefore, we expect to observe higher spreads for loan facilities with non-bank
syndicate member than for loan facilities with bank-only participants.
To evaluate this hypothesis empirically, we estimate equations predicting the interest rate on a
particular loan facility. Because the loans in our sample are floating rate with LIBOR as their index, we
estimate equations predicting the “All-in-Drawn Spread”, which is the spread of the loan facility over
LIBOR plus any annual fees that the borrower must pay the lenders. Our goal is to estimate the
incremental effect of a non-bank institutional investor on the spread, holding other factors that could
affect the spread constant. Therefore, we estimate the following equation:
X member syndicatebankNonβ α spreaddrawninAll
(1)
where X is a vector of covariates that include facility- and firm-specific control variables. The control
variables include the facility amount, the number of participating lenders and the lenders’ past
relationships with the firm, the maturity of the facility, whether the facility is secured or second-lien,
whether the loan has covenants, as well as the borrowing firm’s size, ratio of fixed to total assets, Z-score,
(out of the total 20,031) made to investment grade borrowers, despite the fact that all of our loan facilities are
classified as “leveraged”. 10
“Z-score” is intended to be a negative function of bankruptcy probabilities. It is taken from Altman (1968) and
defined by: Z = 1.2 Working Capital / Total Assets + 1.4 Retained Earnings / Total Assets + 3.3 EBIT / Total Assets
+ 0.6 Market Value of Equity / Book Value of Total Liabilities T4 + 0.999 Sales/ Total Assets.
12
leverage, industry-adjusted ROA, log of the number of analyst following, and total institutional holdings.
We also control for the high-yield spread in the month of facility origination to take into account time-
series variations in market-wide risk premia. Definitions of all variables are provided in Appendix A.
The loan facilities are generally either revolvers or term loans. The term loan facilties are of two
types, Term Loan A facilities, which are often syndicated to banks, or Term Loan B facilities (sometimes
labeled just “Term Loan” in the Dealscan database), which are typically structured for non-bank
institutional investors. Nandy and Shao (2010) document that Term Loan B facilities generally have
higher spreads than Term Loan A facilities, which is consistent with the amortizing nature of Term Loan
A facilities leading to a substantially shorter effective maturity than Term Loan B facilities. For this
reason, it is important to control for differences in type of facility when estimating Equation (1).
A key factor in determining the spread on a loan facility is its default risk. It is possible that
differences in spreads between non-bank and bank-only facilities could reflect the fact that non-bank
facilities tend to be for riskier borrowers (See Table III). To measure the incremental impact of a non-
bank institutional investor on spreads, it is important to control as well as possible for the default risk of
the loan.
About 40% of the firms in our sample has issuer credit ratings at the end of fiscal year prior to the
loan origination. The credit ratings presumably reflect the risk of the issuer as assessed by professionals
around the time the loan is issued. However, relying solely on credit ratings to measure risk necessitates
dropping loan facilities made to firms that do not have ratings. Therefore, we estimate specifications
using issuer credit ratings as a measure of risk for the loan facilities for which credit ratings are available,
as well as equations using the Z-Score and leverage to control for default risk for the larger sample that
includes loan facilities without credit ratings.
We present the OLS coefficient estimates of Equation (1) and the corresponding p-values on the
full sample in Panel A of Table IV, and on the sub-sample with firm credit ratings in Panel B of Table IV.
Each equation includes facility-purpose fixed effects, and year fixed effects, and the reported standard
13
errors are clustered by borrower. When we consider the sample of all facility types, we include facility-
type fixed effects in the equation as well.
Column (1) of Panel A of Table IV presents estimates of Equation (1) using all observations for
which all required data are available (12,346 loan facilities). In this column, the coefficient on the non-
bank syndicate member indicator variable is 56.4, and is statistically significantly different from zero.
This coefficient indicates that holding other things constant, loan facilities with at least one non-bank
syndicate member have spreads that are 56.4 basis points higher than bank-only loan facilities. This
spread difference is relatively large, given the average spread of 249 basis points, so the estimated non-
bank premium equals 22.7% of the total spread.11
The coefficients on the other variables, which control for other factors that potentially affect
spreads, are consistent with the notion that spreads are a function of borrower and loan risk. Larger loan
facilities with more syndicate members, especially when the participants have past relationships with the
borrowers, tend to be less risky and therefore have lower spreads. Secured and second-lien facilities tend
to be more risky, and hence have higher spreads.12
Z-Score has a negative coefficient and leverage a
positive one, suggesting that riskier firms have loan facilities with higher spreads, and profitability in the
form of industry-adjusted ROA, not surprisingly, is associated with lower spreads. We also control for
market-wide risk premium by including the high-yield credit spread, measured as the difference between
the average spread on AAA-rated loan index and the average spread on BB-rated loan index in the month
of loan origination. Not surprisingly, since our sample is relatively risky leveraged loans, the high yield
spread is positively (and statistically-significantly) related to the all-in-drawn spread.
In Column (2) of Panel A of Table IV, we present estimates of Equation (1) for the sub-sample of
revolvers, in Column (3) for all term loan facilities, and in Column (4) for only the Term Loan B facilities.
In each column, the coefficient on the non-bank syndicate member indicator variable is positive and
11
The average spread for the 12,346 facilities that have all required data is 249 basis points (Appendix B). 12
Security by itself lowers the risk of a loan. However, secured loans tend to be issued by younger, riskier firms
with lower cash flows, so the positive relation with spreads likely reflects this additional risk. See Berger and Udell
(1990) and Erel, Julio, Kim, and Weisbach (2012).
14
statistically significantly different from zero. For revolvers, this coefficient implies that loan facilities
having non-bank syndicate members have 47.5 basis-point higher spreads than bank-only loan facilities.
For all term loans pooled together (Column (3)), the premium is 70.3 basis points, and for just Term Loan
B facilities (Column (4)), it is 76.2 basis points. These results imply that there is a positive premium
associated with different types of non-bank loan facilities.
In these equations, we control for default risk based on the borrowing firm’s financial data using
measures such as leverage and the firm’s Z-Score. An alternative way of controlling for risk is to use the
issuer’s credit rating. Since credit ratings are constructed by professionals to measure firms’
comprehensive default risk, it is likely a preferable approach. However, credit ratings are not available for
all firms, so the use of credit ratings is limited to those firms that have them.
Panel B of Table IV presents coefficient estimates of Equation (1) for different credit ratings
(Columns (1) through (4)) and for issuers with no credit ratings (Column (5)). Column (1) contains
estimates for all facilities from issuers with credit ratings, including indicator variables for different credit
rating categories. Columns (2), (3), and (4) present estimates for BBB- and above rated firms, BB-rated
firms, and B-rated firms and below rated firms, respectively. As in Panel A, we include control variables
for facility and firm characteristics, facility-type, facility-purpose and year fixed effects. The reported
standard error estimates are clustered by borrowing firm.
In each column in Panel B of Table IV, the coefficient on the non-bank syndicate member
indicator variable is positive and statistically significantly from zero. For all firms with ratings (Column
(1)), the estimates indicate that there is about a 23.6 basis-point premium for facilities with non-bank
participation. This premium equals 28.1 basis points for firms rated BBB and higher, 21.8 basis points for
BB-rated firms, and 24.1 basis points for B-rated and below firms. In addition, for firms without ratings,
there is an estimated non-bank premium of 78.6 basis points.
3.3. Within-loan estimates.
15
One clear pattern emerging from Table IV is that the estimates of the non-bank premium are
substantially smaller when ratings are used to control for risk when we use borrower-level controls (Panel
A), or for the subsample of firms without ratings in (Panel B, Column (5)). This observation suggests that
the estimated non-bank premium could reflect borrower risk. Credit ratings are themselves imperfect
measures of default risk, since there is variation in risk within rating classes and errors in assigning
ratings to firms. It is unclear if the positive estimated premium for non-bank participation reflects residual
risk not reflected in ratings, or if it reflects an economic premium to attract non-bank institutional lenders
with relatively high required rates of return.
A method of measuring non-bank syndicate member premiums that is unlikely to be affected by
risk or other potential unobserved firm-level heterogeneity comes from the relative pricing of different
facilities within the same loan.13
Since each facility of a multiple facility loan has the same seniority and
covenants, the default risk of facilities and the creditor rights attached to the facilities in the same loan is
essentially the same. Different facilities in the same loan will generally have different maturities and
implicit options from one another that will affect their pricing. However, once these other differences are
controlled for econometrically, the incremental effect of a non-bank participant on the relative pricing of
facilities within a given loan should reflect the impact of non-bank syndicate participation. This approach
will not be affected by differences in risk or some other form of unobservable firm-level heterogeneity
causing a spurious relation between the existence of non-bank institutions in the facility’s syndicate and
the facility’s spread.
Within-loan estimates can help to distinguish between alternative explanations for the non-bank
premium. Any firm-level factor that potentially affects its attractiveness to a lender such as its historical
cash flows, future projects and risks, should affect the spreads on all facilities of the loan similarly. In
contrast, a systematic difference in the relative spread between different facilities that depends on
syndicate composition for a particular facility has to be a function of facility-level rather than firm-level
13
This approach was developed by Ivashina and Sun (2011a) and was recently adopted by Nadauld and Weisbach
(2012).
16
factors. One such possibility would be if different types of facilities have different liquidity and demand.
For example, banks, typically the lead arrangers, hold the Term Loan A or revolver portion of the loan
and sell the Term Loan B portion to non-bank institutional investors. If the lead arranger is worried about
being able to sell a particular Term Loan B facility to banks, it can increase the spread, making the facility
more attractive to non-bank investors such as hedge funds or finance companies. In this situation, the
facilities with non-bank syndicate members are likely to receive unusually high spreads, measured
relative to the Term Loan A or revolver facility of the same loan.
To estimate the incremental effect of a non-bank investor on the differences in spreads between
facilities of a given loan, we estimate the following equation:
Xmembersyndicatenon-bankhasFacilityGapSpread (2)
where X includes differences in facility-specific characteristics such as facility size, the number of
participating lenders, maturity, and whether the facility is secured by collateral, as well as firm-level
characteristics and the high-yield spread. The dependent variable in Equation (2) is the difference
between the spreads of different facilities within the same loan (spread gap). The indicator variable
denoting which facility has a non-bank syndicate member measures the incremental effect of a non-bank
institution on the spread gap, and the control variables are intended to capture other differences between
the facilities that could be related to spreads. We estimate Equation (2) on the sample of loans that have
multiple facilities of the type considered in that specification.
Term Loan B facilities tend to have a longer effective maturity than Term Loan A facilities
because the vast majority of Term Loan A facilities are amortizing while Term Loan B facilities are more
often bullet (see Appendix C). With a longer duration, Term Loan B facilities will have higher spreads
regardless of whether there is a non-bank institution participating in the facility. Consistent with this
argument, Nandy and Shao (2010) document that Term Loan B facilities have higher spreads than Term
17
Loan A facilities or revolvers. The hypothesis that non-bank investors receive premiums relative to
otherwise similar bank-only facilities implies that there should be an additional premium over the
corresponding Term Loan A facility for the Term Loan B facilities in which non-bank institutions invest,
relative to that of an otherwise similar bank-only Term Loan B facility.
Column (1) of Panel A of Table V presents estimates of Equation (2) for the subsample of 246
loans that have both Term Loan A and Term Loan B facilities.14
In this subsample, the non-bank investor
usually participates in the Term Loan B portion of the loan: Of the 246 loans, there were 59 in which non-
bank institutions were syndicate members in the Term Loan B facility and only nine in which the non-
bank institutions participated in the Term Loan A facility. Therefore, we estimate whether the existence
of a non-bank syndicate member in the Term B facility affects the difference in spreads between the two
facilities. The coefficient estimates indicate the presence of a non-bank investor in the Term Loan B
facility increases the difference in spreads between Term Loan B and Term Loan A facilities by 5.3 basis
points. However, this estimated difference in spreads is not statistically significantly different from zero at
conventional levels.
There are 1,608 loans in our sample that have complete data and contain both a Term Loan B
facility and a revolver.15
Of these 1,608 cases, it was more common for the non-bank investor to
participate in the syndicate of the Term Loan B portion of the loan than the revolver portion: non-bank
institutions participated in the syndicates of 131 Term Loan B facilities and only 34 of those of the
revolvers. For this reason, we estimate Equation (2) on this subsample of loans, considering the effect of
the non-bank institution participating in the syndicate of the Term Loan B facility in Column (2). The
dependent variable in these equations is the difference in spreads between the Term Loan B facility and
the revolver. Therefore, the coefficient estimate of 42.3 on the non-bank syndicate member indicator
14
We exclude cases in which non-bank investors are present on both Term Loan A facilities and Term Loan B
facilities. 15
Again, we exclude cases in which non-bank investors are present on both Term Loan B facility and revolver.
18
variable in Column (2) implies that the spread between Term Loan B facilities and revolvers is 42.3 basis-
points higher when a non-bank institution is present in syndicate of the Term Loan B facility.
A potentially cleaner test of the hypothesis that non-bank institutions receive premiums when
investing in syndicated loan facilities comes from cases in which the non-bank invests in one of multiple
facilities of the same type in a particular loan. Our sample contains 217 such facilities, in which non-
banking institutions invested in 106 facilities. In Panel B of Table V, we estimate equations similar to Eqn.
(1) on this sub-sample of loan facilities. In Column (1) we present estimates of Equation (1) without firm-
level controls and in Column (2), we report estimates including these controls. In each equation, the
coefficient estimate on the non-bank syndicate member indicator variable is positive and statistically
significantly different from zero, implying 40.6 and 27.9 basis point premiums to facilities with a non-
bank institutional investor.
Finally, there are 841 cases in which a borrower issues more than one of the same facility type in
the same year, but not necessarily in the same loan. We re-estimate the equation for this subsample in
Columns (3) and (4) of Panel B of Table V. In these equations, the estimated non-bank premium is 33.8
and 34.0 basis points, respectively, each of which is statistically significantly different from zero.
Overall the results from Table V, in which we compare facilities within a given loan, or across
similar facilities from the same borrower within a short period of time, are consistent with the results in
Table IV that are based on comparisons across different facilities. When a non-banking institution
participates in a syndicated loan facility, syndicate members receive a premium on the particular facility
in which non-bank institution invests relative to bank-only facilities. These premiums do not appear to be
a result of unobserved heterogeneity across facilities that is correlated with the facilities’ risk.
4. Types of non-bank institutional syndicate members and spreads
The within-loan results suggest that the premiums to loan facilities that include non-bank
investors in their syndicate occur because of facility-specific and not firm-specific factors. The most
plausible explanation for these premiums is that facilities in which non-bank institutions participate are
19
relatively more difficult to market than bank-only facilities. When the arranger can structure a syndicate
made up entirely of banks, he can charge the borrower a relatively low spread. However, if banks are not
willing to provide the necessary capital, then the arranger will have to charge the borrower a higher
spread to attract capital from investors with higher required rates of return.
This argument has a clear prediction about the premiums we should observer when different types
of non-bank institutional investors are part of the loan facility syndicate. When investors like hedge funds
or private equity funds with high required rates of return invest in a loan facility, it means that the
facility’s arranger had to increase the facility’s spread beyond what would have been necessary if only
banks were the investors. In addition, hedge funds and private fund managers have unusually high
pecuniary incentives, which are likely to motivate them to seek out investments in facilities with
unusually high spreads.16
In contrast, non-bank institutional investors such as insurance companies have
required returns similar to banks, and tend to focus on ensuring that their loan portfolio has the right term
structure and risk profile, rather than seeking out unusual opportunities to achieve abnormal returns.
Therefore, we expect that, controlling for risk, facilities for which hedge and private equity funds are in
the syndicate to have higher spreads than those for which insurance companies invest.
4.1. Abnormal spreads across types of non-bank institutional syndicate members.
In Columns (1) and (2) of Table VI, we re-estimate Equation (1) with the non-bank syndicate
members broken down by institutional type. To do so, we include separate indicator variables for
insurance companies, finance companies, hedge and private equity funds, mutual funds, and other non-
bank investors who could not be classified or are of a type that participated in less than 1% of the full
sample for which all required data are available (about 123 facilities). Column (1) presents estimates of
16
General partners of private equity and hedge funds receive direct incentives through carried interest that usually
equals 20% of profits. In addition, they receive indirect incentives because their performance affects their future
incomes. These indirect incentives are likely to be of similar magnitude as the direct incentives. [See Chung,
Sensoy, Stern and Weisbach (2012) for estimates for private equity funds and Lim, Sensoy and Weisbach (2012) for
estimates for hedge funds.]
20
this equation using all loan facilities, while Column (2) restricts the sample to just those facilities with at
least one non-bank syndicate member.
The coefficient estimates indicate that the premiums vary substantially between types of non-
bank institutional investor. The largest premiums appear to be for facilities with hedge and private equity
fund syndicate members. In Column (1), the coefficient on hedge and private equity funds indicator
variable is 47.2, which implies that loan facilities in which hedge and private equity funds invest have an
abnormal spread of 47.2 basis points relative to the sample of all facilities. When we restrict the sample to
facilities with non-bank syndicate members, the coefficient declines to 29.4, but is still significantly
different from zero. The difference between the two coefficients comes from the implied benchmark from
the sample used; since Column (2) contains only facilities which have non-bank syndicate members, the
sample average spread is higher than for bank-only facilities (controlling for other characteristics).
The other significantly positive premiums occur for facilities with “other” non-bank syndicate
members, which potentially includes hedge funds that we could not classify, and finance companies
(significant only in Column (1)). The coefficients on insurance companies and mutual funds are small and
not statistically significantly different from zero. These results imply that the spread premium varies
across non-bank institutional investor types, and is highest for hedge and private equity funds.
4.2. Within-loan estimates by type of non-bank institutional investor.
The spread premium estimates in Table VI control for the risk of the loan using the borrower’s
financial data. As discussed above, while this approach captures the risk of the facility to some extent, it
does so imperfectly, and it is possible that some of the measured premiums to facilities associated with
non-bank institutional investors could reflect the fact that these investors tend to invest in loans of riskier
firms. To evaluate the extent to which this effect is important, we also estimate spread gap premiums for
different types of non-bank institutional investors using the within-loan approach that compares spreads
of facilities of the same loan.
21
We present the within-loan estimates of the spread gap premiums by type of non-bank
institutional investor in Table VII. Column (1) presents estimates of the difference between the spreads on
the Term Loan B and Term Loan A facilities of the same loan for the loans in our sample that contain
both types of facilities, and Column (2) presents comparable estimates for the difference between Term
Loan B facilities and Revolvers. The positive and statistically significant coefficients on the variable that
indicates that a hedge or private equity fund participated in the syndicate of the Term Loan B facility
implies that when a hedge or private equity fund invests in the Term Loan B facility, the spread gap
between the Term Loan B and Term Loan A facilities appears to be 20 basis points abnormally high.
None of the comparable coefficients indicating that other types of non-bank institutional investors
participated in the syndicate of the Term Loan B facilities are statistically significant. These results
suggest that loan facilities in which hedge and private equity funds invest have higher spreads than
otherwise similar facilities in which other types of institutions invest.
4.3. The size of the non-bank syndicate members’ loan share.
The results so far are consistent with the view that when arrangers are concerned about being able
to raise capital from banks, they increase spreads to attract non-bank institutional investors. An additional
implication of this logic is that arrangers should increase spreads by a larger amount when they require a
greater quantity of capital from the non-bank institutional investors. Therefore, we expect to observe
higher spreads when non-bank stakes in loan facilities are larger.
We evaluate this prediction in Table VIII, using the sub-sample of 3,826 loan facilities for which
DealScan contains data on syndicate member ownership and all other required data are available.17
In
Columns (1) and (2), we include all facilities for which ownership data are available and in Columns (3)
and (4) we restrict the sample to non-bank facilities. Columns (1) and (3) contain the non-bank syndicate
17
We only include facilities for which more than 90% of ownership can be identified. When we further restrict the
sample to facilities having 100% of ownership identified, the sample size decreases to 3,641. The results when we
re-estimate all equations in Table VIII using this smaller sample are similar to those reported in Table VIII.
22
members’ facility share, while Columns (2) and (4) contain an indicator variable that indicates whether
the non-bank institutional investor purchased the largest stake in the loan facility.
The coefficient estimates in Table VIII all suggest that when the non-bank institutional investor
takes a larger stake in the loan facility, spreads are higher. The coefficients on the non-bank syndicate
members’ loan shares in Columns (1) and (3) are positive and statistically significant, as are the
coefficients in Columns (2) and (4) on the indicator variable indicating whether the non-bank institutional
investor has the largest loan share in the loan facility. These results are consistent with the view that
arrangers increase the loan facility’s spread to attract non-bank institutional investors, and the more
capital they have to raise from these investors, the greater the arrangers increase the spread.
4.4. “Dual” holders of both debt and equity.
A number of non-bank investors in syndicated loan facilities also are equity holders in the firm.
Such “dual holding” has become increasingly common in recent years (see Jiang, Li and Shao (2010)).
Presumably, institutional equity holders would utilize their informational or strategic advantage inside the
borrowing firm to improve their other investments including those in the firm’s syndicated loan facilities.
In addition, larger equity ownership implies that the investor will share a larger fraction of the gains
created through a value-increasing loan. We evaluate the extent to which equity ownership influences the
size of the non-bank premium.
To identify whether the non-bank institutional lender held an equity stake in the borrower prior
to the loan origination, we create a list of shareholders of the borrowing company from Thompson Reuters
Institutional Holding Database (13F) for the one-year period leading up to the current loan, as well as the
list of lenders who are participating in the current loan. For example, for a loan originated in April 2000,
we create a list of equity holders using four 13F filings: filings for the quarters that end in June 1999,
September 1999, December 1999, and March 2000, respectively. An institutional investor’s equity stake
is measured as the average of the holdings that appear in these four filings. We focus on the equity stake
held by lenders prior to loan origination because we wish to evaluate the effect of holding an equity
23
position on the loan decision. We then match lender information from DealScan to the institutional
investors in the 13F by the lender’s name, and the lender’s ultimate parent’s name.
In Column (1) of Table IX, we re-estimate Equation (1) including a variable indicating whether
there is a non-bank syndicate member that is a dual holder. The regression includes 314 loan facilities
with participation by a non-bank dual holder. The coefficient on this variable is positive, but not
statistically significantly different from zero. In Column (2) of Table IX, we break up this variable by
type of non-bank institutional investor, and also include the type indicator variables.18
When a hedge fund
or private equity fund is a member of the facility’s syndicate facility’s spread is 29.3 basis points higher.
When a hedge fund or private equity fund also owns at least 0.1% of the borrowing firm’s equity the
facility trades at a premium of 57.7 basis points (= 29.29 + 28.46), which is significant at the 1% level.
This finding suggests that, especially when they are equity holders, hedge and private equity funds can be
viewed as lenders of last resort, and will lend to firms but only at a large premium.
5. The Uses of Funds
Thus far the results indicate that arrangers increase the loan facility spread when it is necessary to
attract non-bank institutional investors who have a relatively high required rates of return. An additional
implication of this result is that firms are likely to seek out non-bank institutional investors as participants
in the loan facility if the hard-to-fund facility also is particularly important to the firm. The idea is that, if
it is difficult to raise a specific facility, firms will only be willing to pay a higher spread to get a non-bank
institutional investor to participate if the capital being raised has a valuable use that cannot be delayed to a
point in time in the future when it could be unnecessary to pay an additional premium to acquire capital.
This argument predicts that firms borrowing at abnormally high spreads will spend a higher proportion of
the raised cash relatively quickly, rather than saving it as additional cash.
18
The regression includes 17 facilities with insurance company dual holders, 105 facilities with finance company
dual holders, 47 facilities with hedge fund and private equity fund dual holders, 215 facilities with mutual fund dual
holders, and 23 facilities with other types of non-bank institutional dual holders.
24
To test this prediction, we estimate models similar to those in Kim and Weisbach (2008) that
predict the change in cash holdings in a particular time period following the capital raising. The idea is
that an abnormally high spread should lead firms to spend the capital they raise more quickly, so that they
should save less of it in cash. The measure of abnormal spread we use is the residual from the equation
presented in Column (1) of Panel A of Table IV, but excluding the non-bank syndicate member indicator.
We estimate whether it predicts the change in cash holdings subsequently, using the following equation:
εQtr-YearθassetstotallnβResidual Spreadβ
Residual Spread1assetstotal
proceedloanlnβ
1assetstotal
proceedloanlnββ1
assetstotal
cashcashln
2007Q4
1997Q1i
i043
0
2
0
10
0
0t
(3)
where time is indexed by quarters subsequent to the loan issuance quarter.
We present estimates of this equation in Panel A of Table X. Each row of this table represents a
regression predicting the change in normalized cash as over a specified time period. The top row predicts
the change in cash during the first quarter following the capital raising, the second, the change in cash
during the four quarters following the capital raise, and the third row the change in cash during the eight
quarters following the capital raising. In the first row, the positive coefficient on β1 means that a
significant portion of capital raised is used to increase cash holdings in the first quarter following a loan
facility. However, the negative coefficient on β2 implies that the fraction of loan facility used to increase
cash holdings is smaller for firms that raised capital at higher spreads, consistent with the view that when
spreads are higher, firms are more likely to spend a higher fraction of the money raised and save less as
cash.
We also test the prediction that the effect of higher loan facility pricing on the uses of funds
should be larger when a non-bank investor is included in the syndicate. To perform this test, we estimate
25
Equation (3) separately for non-bank participated facilities and bank-only facilities. The results using this
specification are presented in Panel B and C of Table X. The coefficients on the spread residual (β2) are
all negative for non-bank participated facilities, while they are mostly positive for bank-only facilities.
The differences in β2 are statistically significant at the 10% level for first quarter and for the four quarters
subsequent to the capital raising. These results suggest that firms are willing to pay a higher spread to
attract non-bank institutional investors when raising capital is particularly important to the firm.
6. Conclusion
Participation by non-bank institutions has become a major part of the syndicated loan market. In
our sample of 20,031 “leveraged” loan facilities originated between 1997 and 2007 from the DealScan
database, 6,279 facilities have at least one a non-bank institution syndicate member. Some of these non-
bank institutions have substantially higher required returns than banks, yet both banks and non-bank
institutions invest in the same loan facilities. One explanation for this phenomenon is that loan arrangers
approach non-bank institutional investors when they cannot fill the syndicate with banks, and
consequently have to offer a higher spread to attract non-bank institutional investors.
We estimate the abnormal spread that a non-bank institutional investor receives by comparing
spreads on loan facilities with non-bank institutional investors to those on observationally equivalent
facilities that do not have a non-bank institutional investor. Our estimates indicate, holding all else equal,
that loan facilities with a non-bank syndicate member receive a higher spread than otherwise similar
facilities with bank-only syndicates. The positive spread is statistically and economically significant for
revolvers as well as term loan facilities and for loan facilities to borrowers of different credit ratings as
well as unrated borrowers.
It is possible that the presence of a non-bank institutional investor is correlated with other,
potentially unobservable factors related to the loan facility’s spread, which could drive the non-bank
premiums. For example, it is possible that the risk of the firms in which non-bank institutions tend to
invest tends to be higher than is reflected in their ratings. To address this possibility, we use a “within-
26
loan” estimation approach that compares differences in spreads across facilities of the same loan. Since
different facilities of the same loan share the same underlying risk and have the same seniority,
unobservable differences in risk cannot explain differences in spreads of facilities of the same loan.
Because factors such as maturity and implicit options affect the spreads of different types of
facilities, we test whether the existence of a non-bank institutional investor affects the relative difference
in spreads, holding other factors constant. Our results suggest that in a loan with both a Term Loan B
facility and a revolver, if a non-bank institution invests in a Term Loan B facility, the spread between the
two is higher than would be expected without non-bank participation. In the subsample of non-bank loans
that have multiple tranches of the same type, the facilities with non-bank institutional investor
participation have higher spreads than the facilities without non-bank participation. These results are not
consistent with the view that non-bank premiums reflect unobservable risk. Instead, they suggest that non-
bank institutional investors can be viewed as lenders of last resort, and receive higher spreads because
they are willing to provide capital at times when banks are not.
Our results suggest that there are substantial differences in premiums going to different types of
non-bank institutional investors. When private equity and hedge funds are non-bank investors, they
receive a 47.2 basis-point premium over other loan facilities. In contrast, other types of non-banks
institutional investors such as insurance companies or mutual funds receive essentially no abnormal
premium at all. In addition, abnormal spreads are higher when the hedge or private equity funds have
equity positions in the firm, and when they purchase a larger share of the loan facility.
Non-bank institutional investors, especially private equity and hedge funds, have become
important lenders to corporations through their role in the syndicated loan market. The evidence in this
paper suggests that the non-bank institutions obtain higher interest rates than other investors. This spread
premium appears to be due to the circumstances under which capital is provided rather than unobserved
borrower risk.
As debt markets mature, it seems evident that non-traditional players will provide capital to a
larger degree than has been true historically. Our results suggest that non-bank institutions provide capital
27
when capital raising is important to firms and receive a premium for providing the financing. Why is it
optimal to have different types of investors providing the capital for the same loans? To what extent does
borrower performance depend on the provider of capital? Is there important variation across banks that
leads some to be more prone to co-invest with hedge funds and private equity funds in loans with higher
spreads? Understanding the answers to these and related questions would be a useful direction for future
research.
28
References
Altman, E. I. (1968) “Financial Ratios, Discriminant Analysis, and the Prediction of Corporate
Bankruptcy,” Journal of Finance 23, 589–609.
Berger, Allen N. and Gregory F. Udell, (1990) “Collateral, loan quality, and bank risk,” Journal of
Monetary Economics 25, 21-42.
Brophy, David J., Paige P. Ouimet, and Clemens Sialm (2009) “Hedge Funds as Investors of Last Resort?”
Review of Financial Studies, 22, 541-574.
Chava, S., and M. R. Roberts. (2008) “How does financing impact investment? The role of debt
covenants,” The Journal of Finance 63 (5): 2085-121.
Chung, Ji-Woong, Berk A. Sensoy, Léa H. Stern, and Michael S. Weisbach (2012), “Pay for
Performance from Future Fund Flows: The Case of Private Equity,” Review of Financial Studies,
forthcoming.
Erel, Isil, Brandon Julio, Woojin Kim and Michael S. Weisbach (2012) “Macroeconomic Conditions and
Capital Raising,” Review of Financial Studies, 25, 341-376.
Ivashina, Victoria, and Zheng Sun. (2011a) “Institutional Demand Pressure and the Cost of Corporate
Loans,” Journal of Financial Economics 99, 500-522.
Ivashina, Victoria, and Zheng Sun (2011b) “Institutional Stock Trading on Loan Market Information,”
Journal of Financial Economics, 100, 284-303.
Jiang, Wei, Kai Li, and Peng Shao (2010) “When shareholders are creditors: Effects of the simultaneous
holding of equity and debt by non-commercial banking institutions,” Review of Financial Studies 23,
3595-3637.
Kim, Woojin and Michael S. Weisbach (2008) “Motivations for Public Equity Offers: An International
Perspective,” Journal of Financial Economics, 87, 281-307.
Lim, Jongha, Berk A. Sensoy, and Michael S. Weisbach (2012) “Total Incentives of Hedge Fund
Managers,” Research in Progress, Ohio State University.
Massoud, N., D. Nandy, A. Saunders, and K. Song (2010) “Do hedge funds trade on private information?
Evidence from syndicated lending and short selling.” Journal of Financial Economics.
Nadauld, Taylor D. and Michael S. Weisbach (2012) “Did Securitization Affect the Cost of Corporate
Debt?” Journal of Financial Economics, 105, 332-352.
Nandy, D., and P. Shao (2010) “Institutional investment in syndicated loans.” Working Paper, Brandeis
University.
29
Table I. Trends in non-bank institutional participation in leveraged loan facilities
This table presents the trends in the distribution of loan facility originations during 1997-2007 by number (Panel A) and dollar value (Panel B). Column (1)
reports the total number (value) of all leveraged loan facilities from the DealScan database. Column (2) reports the total number of loan facilities in which only
commercial or investment banks participated. Columns (3) and reports the total number (value) of loan facilities in which at least one non-bank institution is a
member of the facility’s syndicate. Columns (4) – (10) report the total number (value) of loan facilities by type of institutional syndicate member. The sum of
columns (4) to (10) do not add to the total number (value) in column (1) because than one type of institution can participate in the syndicate.
Type of institutional syndicate member
Year of
Origination
All facilities
Bank-only
syndicate
Non-bank
syndicate
member
Commercial
Bank
Investment
Bank
Insurance
Company
Finance
Company
HF/PE Mutual
Fund
Other
Panel A: Number of loan facilities
1997 2,706 2,267 439 2,527 708 79 331 146 128 42
1998 2,264 1,783 481 2,111 598 80 357 154 111 73
1999 1,915 1,424 491 1,791 606 82 394 192 137 73
2000 1,780 1,324 456 1,656 524 48 349 174 93 68
2001 1,777 1,328 449 1,628 603 57 317 192 67 72
2002 1,800 1,178 622 1,653 623 73 442 280 85 79
2003 1,778 982 796 1,644 745 120 565 420 106 106
2004 1,818 941 877 1,643 879 90 638 484 142 119
2005 1,626 921 705 1,463 920 30 539 304 71 63
2006 1,365 805 560 1,198 806 21 406 228 43 27
2007 1,202 799 403 1,087 711 18 265 180 27 29
Total 20,031 13,752 6,279 18,401 7,723 698 4,603 2,754 1,010 751
Panel B: Value of loan facilities (in $ billions)
1997 298 242 57 294 192 16 45 20 27 8
1998 258 182 76 254 158 21 58 28 24 14
1999 255 182 73 252 167 20 57 31 32 15
2000 237 159 78 232 139 13 61 32 27 17
2001 256 180 76 246 177 14 65 30 15 13
2002 231 132 99 222 162 17 79 45 20 16
2003 257 115 141 247 181 36 115 83 30 25
2004 327 155 172 307 251 22 138 95 38 22
2005 349 196 153 326 278 10 125 60 22 15
2006 342 210 132 318 272 6 107 40 18 6
2007 345 235 110 328 285 7 85 39 12 6
Total 3,160 1,990 1,170 3,026 2,262 182 935 501 266 156
30
Table II. Selected facility and lender characteristics
This table presents sample averages of selected facility (Panel A) and lender characteristics (Panel B, C, and D).
Averages are reported for the full sample of facilities and for the sub-samples of Revolers, Term Loan A facilities,
and Term Loan B facilities. Panel B, C and D include only loan facilities for which more than 90% of loan shares
can be identified (5,624 facilities). The sample of leveraged loan facilities is from the DealScan database, originated
between 1997 and 2007.All continuous variables are winsorized at 1% and 99% level. Definitions of the variables
are provided in Appendix A.
Facility Type
All Facilities Revolver Term Loan A Term Loan B
Panel A: Facility Characteristics
N 20,031 12,421 956 6,654
Non-bank participated 0.313 0.249 0.482 0.409
Facility amount ($M) 158.0 153.0 174.0 164.0
Number of participating lenders 6.110 5.836 9.476 6.136
% of participating lenders with past relationship 0.304 0.302 0.338 0.303
Maturity 46.87 39.91 58.19 58.22
Secured Indicator 0.701 0.666 0.719 0.763
Second-lien Indicator 0.014 0.000 0.005 0.041
Covenants Indicator 0.756 0.762 0.715 0.750
Syndicated Indicator 0.876 0.862 0.978 0.886
All-in-drawn spread (bps) 256.6 230.1 271.4 305.8
Panel B: Participation by syndicate member type - conditional on having loan share information
All bank 5,333 3,938 134 1,261
Commercial bank 5,274 3,910 133 1,231
Investment bank 1,535 1,127 83 325
All non-bank 1,282 784 61 437
Insurance company 119 27 6 86
Finance company 981 619 53 309
HF/PE 463 246 19 198
Mutual fund 184 47 6 131
Other lenders 125 44 3 78
Panel C Average loan share - conditional on participation (%)
All bank 94.7 96.0 91.8 90.9
Commercial bank 89.8 91.2 80.0 86.3
Investment bank 20.8 19.2 20.0 26.4
All non-bank 44.2 42.2 24.9 50.6
Insurance company 13.6 13.2 6.9 14.2
Finance company 35.2 35.7 19.3 36.9
HF/PE 31.8 35.6 19.0 28.3
Mutual fund 10.8 6.8 17.2 15.2
Other lenders 31.9 18.7 20.9 24.7
Panel D % of syndicate member type as largest lender - conditional on participation
All bank 96.5 97.2 97.0 94.4
Commercial bank 94.2 95.5 89.5 90.3
Investment bank 28.9 26.4 32.5 36.6
All non-bank 46.4 48.5 19.7 46.5
Insurance company 13.4 18.5 16.7 11.6
Finance company 40.9 44.3 15.1 38.5
HF/PE 29.6 37.0 15.8 21.7
Mutual fund 15.2 17.0 0.0 15.3
Other lenders 23.2 27.3 33.3 20.5
31
Table III. Differences in attributes of non-bank facilities and bank-only facilities
This table shows the differences in various attributes between non-bank loan facilities and bank-only loan facilities
in our sample. A non-bank facility is a facility for which there is at least one non-bank institution in the syndicate.
Definitions of the variables are provided in Appendix A. The total number of loan facilities in our sample with a full
set of data is 20,031 of which 13,752 are bank-only facilities and 6,279 are non-bank facilities. Panel A, B, C and D
present the differences in the type of facility purchased, issuer credit rating, facility characteristics and borrowing
firm characteristics, respectively.
Non-bank
facilities
(1)
Bank-only
facilities
(2)
Difference
(1) - (2)
N Mean N Mean Diff. (t-value)
Panel A. Facility type
% of Revolver 6,279 49.3 13,752 67.8 -18.5 (-25.36)***
% of Term A 6,279 7.3
13,752 3.6 3.7 (11.56)***
% of Term B 6,279 43.3 13,752 28.6 14.7 (20.73)***
Panel B. S&P Issuer credit rating
% of having a credit rating 6,279 52.6 13,752 34.8 17.7 (24.06)***
Conditional on having a credit rating:
% of BBB and above 3,301 5.1
4,792 19.4 -14.3 (-18.79)***
% of BB 3,301 38.9
4,792 44.6 -5.7 (-5.07)***
% of B and below 3,301 56.0 4,792 36.0 19.9 (18.08)***
Panel C. Facility characteristics
Facility amount ($M) 6,279 186.0 13,752 145.0 41.0 (10.81)***
Number of participating lenders 6,279 8.882
13,752 4.844 4.038 (37.84)***
% of participating lenders with past relationship 6,279 27.19
13,752 31.88 -0.047 (-8.31)***
Maturity (months) 6,279 54.40
13,752 43.43 10.97 (30.91)***
Secured indicator 6,279 0.791
13,752 0.659 0.131 (18.99)***
Second-lien indicator 6,279 0.025
13,752 0.009 0.016 (9.11)***
Covenants indicator 6,279 0.753
13,752 0.757 -0.004 (-0.63)
Syndicated facility indicator 6,279 0.944 13,752 0.844 0.100 (20.0)***
Panel D. Borrowing firm characteristics
Total assets ($M) 5,519 1,975.9 12,910 1,602.2 373.8 (5.48)***
Fixed assets/total assets 5,437 0.319
12,678 0.320 0.000 (-0.11)
Z-score 4,032 2.182
10,144 3.546 -1.363 (-19.95)***
Leverage 4,957 0.781
11,979 0.610 0.171 (26.73)***
Industry-adjusted ROA 5,428 -0.110
12,739 -0.074 -0.036 (-10.34)***
Number of analysts following 6,279 3.289
13,752 3.870 -0.581 (-8.03)***
Institutional holdings 6,279 0.290 13,752 0.332 -0.042 (-8.42)***
32
Table IV. Do loan facilities with non-bank syndicate members have higher or lower spreads?
This table presents the OLS regression coefficient estimates of Equation (1) and corresponding p-values. Definitions of all variables are provided in Appendix A.
The dependent variable is the all-in-drawn loan spread over LIBOR in basis points, and the analysis is conducted at the loan facility level. Panel A reports the
results for regressions estimated by type of facility. Panel B reports the results for regression estimated by credit rating groupings for the sub-sample of firms
with S&P issuer credit ratings. The number of loan facilities for which all required data are not missing is 12,346. All specifications include facility-purpose
fixed effects and year fixed effects. The specifications in Column (1) of Panel A and all columns in Panel B additionally include facility-type fixed effects,
because they consider the full sample of all facility types. Standard errors are clustered at the firm level. ***, **, * correspond to statistical significance at the 1%,
5%, and 10% level, respectively.
Panel A: By Facility type
All Facility Types
Revolvers
All Term Loan
facilities
Term Loan B facilities
Dependent Var.= All-in-drawn spread (1) (2) (3) (4)
coef. (p-val) coef. (p-val) coef. (p-val) coef. (p-val)
Non-bank syndicate member 56.414*** (0.000) 47.460*** (0.000) 70.287*** (0.000) 76.155*** (0.000)
Log(Facility amount) -13.840*** (0.000) -22.128*** (0.000) -3.205 (0.147) -4.317* (0.067)
Log(Number of participating lenders) -13.755*** (0.000) -5.460*** (0.001) -25.956*** (0.000) -26.447*** (0.000)
% of participating lenders with past relationship -0.452 (0.885) 3.817 (0.195) -7.963 (0.160) -6.920 (0.246)
Log(Maturity) -16.118*** (0.000) -9.976*** (0.000) -27.457*** (0.000) -25.811*** (0.000)
Secured Indicator 43.085*** (0.000) 46.731*** (0.000) 35.420*** (0.000) 35.137*** (0.000)
Second-lien Indicator 303.904*** (0.000) 252.803*** (0.001) 300.028*** (0.000) 295.064*** (0.000)
Covenants Indicator -13.517*** (0.000) -15.885*** (0.000) -11.107 (0.119) -8.707 (0.228)
Syndicated facility Indicator -9.354** (0.030) -6.170 (0.124) -9.568 (0.290) -11.823 (0.201)
Log(Total assets) -0.656 (0.655) 0.133 (0.927) -1.684 (0.478) -1.174 (0.642)
Fixed assets/total assets 1.628 (0.759) 0.405 (0.936) -0.516 (0.957) -3.377 (0.738)
Z-score -2.361*** (0.000) -2.328*** (0.000) -2.257*** (0.004) -2.349*** (0.005)
Leverage 31.058*** (0.000) 34.549*** (0.000) 21.877*** (0.009) 24.194*** (0.006)
Industry-adjusted ROA -81.533*** (0.000) -77.107*** (0.000) -88.939*** (0.000) -87.586*** (0.000)
Log(Number of analyst following) -5.872*** (0.005) -4.087** (0.036) -9.207** (0.011) -7.835** (0.035)
Institutional holdings -16.004*** (0.001) -20.306*** (0.000) -10.433 (0.207) -8.201 (0.346)
High-yield spread 0.089*** (0.000) 0.063*** (0.002) 0.140*** (0.001) 0.142*** (0.001)
Term A facility
-8.866** (0.028)
Number of observations 12,346 8,065 4,281 3,752
Adjusted R2 0.464 0.451 0.426 0.421
33
Panel B: By issuer credit rating
All ratings BBB-rated and Above BB-rated B-rated and Below No credit rating
Dependent Var.= All-in-drawn spread (1) (2) (3) (4) (5)
coef. (p-val) coef. (p-val) coef. (p-val) coef. (p-val) coef. (p-val)
Non-bank syndicate member 23.554*** (0.000) 28.147** (0.010) 21.752*** (0.000) 24.109*** (0.000) 78.639*** (0.000)
Log(Facility amount) -15.929*** (0.000) -6.207 (0.137) -12.236*** (0.000) -20.405*** (0.000) -13.524*** (0.000)
Log(Number of participating lenders) -12.451*** (0.000) -13.534*** (0.003) -15.397*** (0.000) -9.116*** (0.007) -10.800*** (0.000)
% of participating lenders with past relationship -9.734** (0.015) -1.247 (0.899) -8.782** (0.050) -14.546** (0.041) 0.021 (0.995)
Log(Maturity) -9.979*** (0.001) 2.268 (0.637) 4.085 (0.305) -32.726*** (0.000) -16.284*** (0.000)
Secured Indicator 36.413*** (0.000) 49.944*** (0.000) 39.559*** (0.000) 21.736** (0.012) 36.267*** (0.000)
Second-lien Indicator 265.633*** (0.000) 305.210*** (0.000) 227.651*** (0.000) 264.135*** (0.000) 314.732*** (0.000)
Covenants Indicator -4.086 (0.380) -15.352* (0.050) -7.288 (0.189) 9.845 (0.305) -17.065*** (0.000)
Syndicated facility Indicator 38.407*** (0.007) 45.672 (0.126) 39.432 (0.198) 41.564** (0.020) -9.927** (0.020)
Log(Total assets) 7.016*** (0.000) 11.181*** (0.008) 5.199** (0.027) 7.329*** (0.007) -4.909*** (0.005)
Fixed assets/total assets 9.326 (0.129) 21.570 (0.122) -0.966 (0.899) 16.117 (0.130) 2.417 (0.700)
Leverage 18.067*** (0.001) -7.818 (0.619) 20.478*** (0.006) 18.923** (0.013) 23.077*** (0.000)
Industry-adjusted ROA -66.103*** (0.000) -45.137 (0.423) -
114.122*** (0.000) -47.351*** (0.006) -77.780*** (0.000)
Log(Number of analyst following) -4.835** (0.029) -6.550 (0.177) -6.227** (0.023) -3.085 (0.396) -3.646 (0.127)
Institutional holdings -8.321* (0.087) 5.589 (0.590) -9.945 (0.116) -10.766 (0.205) -17.551*** (0.006)
High-yield spread 0.163*** (0.000) 0.109 (0.169) 0.181*** (0.000) 0.163*** (0.001) 0.050* (0.068)
BB-rated 32.023*** (0.000)
B-rated and below 83.975*** (0.000)
Number of observations 6,879 926 2,984 2,969 8,646
Adjusted R2 0.540 0.382 0.485 0.417 0.434
34
Table V. Is the non-bank premium driven by unobservable heterogeneity across firms?
Panel A presents the OLS regression coefficient estimates of Equation (2) and corresponding p-values on the sample
of loans that have multiple facilities. Definitions of all variables are provided in Appendix A. The dependent
variable is the spread gap between the all-in-drawn spreads of different facilities within the same loan in basis points.
The indicator variable denoting the non-bank facility measures the incremental effect on spread gap of the non-bank
institution participating in the syndicate of the loan facility, and the control variables are intended to capture other
differences between the facilities. Column (1) of Panel A presents estimates for the sub-sample of 246 facility pairs
that have both Term Loan A and Term Loan B facilities, and Columns (2) and (3) present estimates for the sub-
sample of 1,608 facility pairs that have both a Term Loan B facility and a Revolver. Panel B presents the OLS
regression coefficient estimates of Equation (1) and corresponding p-values on the sample of non-bank loan
facilities and the matched bank-only loan facilities of the same facility type. Column (1) employs 106 loans (217
facilities) that have both a non-bank facility and a bank-only facility of the same facility type within the same loan.
Column (3) considers 420 non-bank loan facilities and 421 matched bank-only loan facilities of the same facility
type issued by the same borrower in the same year, but not necessarily in the same loan. Number of observations
drops in Column (2) and (4), as we include firm-level control variables. All specifications include facility-purpose
fixed effects and year fixed effects. All regressions in Panel B additionally include facility type fixed effects.
Standard errors are clustered at the firm level. ***, **, * correspond to statistical significance at the 1%, 5%, and 10%
level, respectively.
Panel A. Within-deal spread gap between facilities
(Term B facility – Term A facility) spread in the same loan
(Term B facility - Revolver) spread in the same loan
Dependent Var.: Within-loan spread gap between facilities (1) (2)
coef. (p-val) coef. (p-val)
Non-bank syndicate member in Term B facility 5.298 (0.546) 42.291*** (0.000)
Differences in Log(Facility amount) -0.788 (0.889) -7.177*** (0.000)
Differences in Log(Number of participating lenders) 3.473 (0.525) -10.178** (0.018)
Differences in % of participating lenders with past relationship 2.824 (0.898) 13.057 (0.320)
Differences in Log(Maturity) -24.965 (0.191) -1.287 (0.699)
Differences in Security -47.639 (0.233) -12.709 (0.246)
Differences in Second-lien 246.379*** (0.000) 255.352*** (0.000)
High-yield spread -0.102 (0.365) 0.039 (0.240)
Firm-level controls Yes Yes
Number of observations 246 1,608
# Non-bank syndicate member in Term B facility 59 131
Adjusted R2 0.234 0.196
35
Panel B. Differences in spreads between non-bank facilities and matching bank-only facilities
Matching Same facility type
within the same oan
Same facility type to the same borrower in the
same calendar year
Dependent Var.: All-in-drawn spread (1) (2) (3) (4)
coef. (p-val) coef. (p-val) coef. (p-val) coef. (p-val)
Non-bank syndicate member 40.618*** (0.001) 27.851** (0.048)
33.800*** (0.000) 34.021*** (0.005)
Log(Facility amount) -13.308*** (0.009) -6.867 (0.254)
-22.525*** (0.000) -14.430* (0.056)
Log(Number of participating lenders) -28.216*** (0.000) -21.715*** (0.005)
-25.112*** (0.000) -25.901*** (0.004)
% of participating lenders with past
relationship -37.405** (0.028) -30.042 (0.156)
-28.295** (0.031) -14.573 (0.403)
Log(Maturity) 17.845 (0.184) 6.948 (0.728)
-9.238 (0.279) -13.282 (0.206)
Secured Indicator 40.528*** (0.007) 39.763** (0.021)
30.491** (0.022) 30.709* (0.063)
Second-lien Indicator 329.106*** (0.000) 256.792*** (0.000)
278.000*** (0.000) 299.089*** (0.000)
Covenants Indicator -9.758 (0.781) -59.703* (0.090)
2.183 (0.889) 16.528 (0.408)
Syndicated facility Indicator -19.072 (0.576) -63.791* (0.079)
36.358 (0.340) -25.584 (0.602)
High-yield spread 0.110 (0.379) 0.105 (0.432)
0.088 (0.358) 0.078 (0.571)
Firm-level controls No Yes
No Yes
Number of observations 217 130 841 496
# Non-bank facilities 106 64
420 243
Adjusted R2 0.715 0.763 0.418 0.441
36
Table VI. Does the type of non-bank syndicate member affect the pricing of the loan facility?
This table presents the OLS regression coefficient estimates of Equation (1) and corresponding p-values, with the
non-bank institutions broken down by the type of institution. Definitions of all variables are provided in Appendix A.
The dependent variable is the all-in-drawn loan spread over LIBOR in basis points, and the analysis is conducted at
the loan facility level. Column (1) uses the full sample of loan facilities and column (2) uses the sub-sample of non-
bank loan facilities. All specifications include facility-type fixed effects, facility-purpose fixed effects, and year
fixed effects. Standard errors are clustered at the firm level. ***, **, * correspond to statistical significance at the
1%, 5%, and 10% level, respectively.
All loan facilities Non-bank loan facilities
Dependent Var.= All-in-drawn spread (1) (2)
coef. (p-val) coef. (p-val)
Insurance company syndicate member -3.544 (0.655) 0.463 (0.947)
Finance company syndicate member 30.906*** (0.000) 3.587 (0.539)
HF/PE syndicate member 47.169*** (0.000) 29.399*** (0.000)
MF syndicate member 0.709 (0.906) 6.587 (0.259)
Other non-bank institutional syndicate member 32.746*** (0.000) 27.337*** (0.000)
Log(Facility amount) -13.642*** (0.000) -26.093*** (0.000)
Log(Number of participating lenders) -15.340*** (0.000) -21.939*** (0.000)
% of participating lenders with past relationship 0.957 (0.760) -16.211** (0.024)
Log(Maturity) -14.975*** (0.000) -19.765*** (0.000)
Secured Indicator 44.047*** (0.000) 18.138** (0.020)
Second-lien Indicator 301.199*** (0.000) 304.136*** (0.000)
Covenants Indicator -14.155*** (0.000) 0.393 (0.962)
Syndicated facility Indicator -8.852** (0.042) -16.253 (0.249)
Log(Total assets) -0.522 (0.725) -1.247 (0.642)
Fixed assets/total assets -0.072 (0.989) 18.508* (0.094)
Z-score -2.441*** (0.000) -3.017*** (0.006)
Leverage 31.533*** (0.000) 10.928 (0.206)
Industry-adjusted ROA -82.095*** (0.000) -66.242*** (0.000)
Log(Number of analyst following) -6.230*** (0.003) -7.174* (0.067)
Institutional holdings -15.433*** (0.002) -12.937 (0.134)
High-yield spread 0.092*** (0.000) 0.089** (0.042)
Number of observations 12,346 3,460
Adjusted R2 0.461 0.500
37
Table VII. Does the type of non-bank syndicate member affect the pricing of the loan facility? – Within-loan
analysis
This table presents the OLS regression coefficient estimates of Equation (2) and corresponding p-values on the
sample of loans that have multiple facilities. Definitions of all variables are provided in Appendix A. The dependent
variable is the spread gap between the all-in-drawn spreads of different facilities within the same loan in basis points.
The indicator variable indicating the type of non-bank syndicate member measures the incremental effect on spread
gap of that type of non-bank institution participating in the syndicate of the loan facility, and the control variables
are intended to capture other differences between the facilities. Column (1) presents estimates for the sub-sample of
252 facility pairs that have both Term Loan A and Term Loan B facilities, and Column (2) estimates for the sub-
sample of 1,615 facility pairs that have both a Term Loan B facility and a Revolver. All specifications include firm-
level control variables, facility-purpose fixed effects and year fixed effects. Standard errors are clustered at the firm
level. ***, **, * correspond to statistical significance at the 1%, 5%, and 10% level, respectively.
Term loan B facility
spread – Term loan A
facility spread
Term loan B facility
spread - Revolver spread
Dependent Var.: Within-loan spread gap between facilities (1) (2)
coef. (p-val) coef. (p-val)
Insurance company syndicate member in Term loan B facility 19.298 (0.142) 8.263 (0.465)
Finance company syndicate member in Term loan B facility -18.905* (0.096) -8.551 (0.530)
HF/PE syndicate member in Term loan B facility 20.049** (0.049) 66.224*** (0.005)
Mutual fund syndicate member in Term loan B in facility -6.489 (0.436) -24.694 (0.200)
Other non-bank syndicate member in Term loan B facility 1.818 (0.864) -0.859 (0.953)
Differences in Log(Facility amount) 0.543 (0.925) -7.187*** (0.000)
Differences in Log(Number of participating lenders) 4.695 (0.410) -10.768** (0.018)
Differences in % of participating lenders with past relationship 13.948 (0.523) 6.886 (0.600)
Differences in Log(Maturity) -25.348 (0.181) -0.891 (0.786)
Differences in Security -48.607 (0.227) -14.581 (0.171)
Differences in Second-lien 252.945*** (0.000) 271.356*** (0.000)
High-yield spread -0.114 (0.303) 0.032 (0.334)
Constant 123.074*** (0.009) 24.271** (0.036)
Firm-level controls Yes Yes
Adjusted R2 0.199 0.222
Number of observations 252 1,615
# Insurance company syndicate members in Term loan B facility 32 54
# Finance company syndicate members in Term loan B facility 45 87
# HF/PE syndicate members in Term loan B facility 53 110
# Mutual fund syndicate members in Term loan B facility 55 87
# Other non-bank syndicate members in Term loan B facility 30 59
38
Table VIII. Does the size of non-bank institutional syndicate members’ loan facility share affect the pricing of the loan facility?
This table presents the OLS regression coefficient estimates of Equation (1) and corresponding p-values. Equation (1) is augmented to include a measure of the non-
bank syndicate members’ share in the loan facility (Columns (1) and (3)) and whether a non-bank syndicate member is the largest lender (Columns (2) and (4)).
Definitions of all variables are provided in Appendix A. The dependent variable is the all-in-drawn loan spread over LIBOR in basis points, and the analysis is
conducted at the loan facility level. Columns (1) and (2) use the full sample of loan facilities and Columns (3) and (4) use the sub-sample of non-bank loan facilities.
All specifications include facility-type fixed effects, facility-purpose fixed effects, and year fixed effects. Standard errors are clustered at the firm level. ***, **, *
correspond to statistical significance at the 1%, 5%, and 10% level, respectively.
All loan facilities Non-bank facilities
Dependent Var.= All-in-drawn spread (1) (2) (3) (4)
coef. (p-val) coef. (p-val) coef. (p-val) coef. (p-val)
Non-bank syndicate members’ loan facility share 96.360*** (0.000)
95.442*** (0.000)
Non-bank syndicate member is the largest lender
44.121*** (0.000)
21.531* (0.051)
Non-bank syndicate member 21.610*** (0.002) 43.891*** (0.000)
Log(Facility amount) -16.257*** (0.000) -16.048*** (0.000) -11.581** (0.032) -12.961** (0.019)
Log(Number of participating lenders) -4.573 (0.208) -7.256** (0.044) 1.179 (0.904) -11.279 (0.204)
% of participating lenders with past relationship -4.232 (0.387) -5.280 (0.281) -12.195 (0.392) -19.783 (0.169)
Log(Maturity) -18.358*** (0.000) -18.268*** (0.000) -30.948*** (0.005) -28.810*** (0.008)
Secured Indicator 45.898*** (0.000) 44.991*** (0.000) 36.537*** (0.006) 39.759*** (0.003)
Second-lien Indicator 284.408*** (0.000) 299.114*** (0.000) 289.119*** (0.000) 309.414*** (0.000)
Covenants Indicator -8.334 (0.279) -9.435 (0.224) -27.690 (0.157) -28.110 (0.149)
Syndicated facility Indicator -21.655*** (0.001) -21.290*** (0.001) -22.460 (0.344) -41.793* (0.063)
Log(Total assets) -2.852 (0.250) -3.081 (0.215) -9.943* (0.100) -9.474 (0.114)
Fixed assets/total assets -6.842 (0.388) -6.051 (0.449) 2.849 (0.890) 2.099 (0.921)
Z-score -2.438*** (0.000) -2.573*** (0.000) -5.299*** (0.008) -5.712*** (0.004)
Leverage 32.343*** (0.000) 33.998*** (0.000) 13.609 (0.453) 18.148 (0.319)
Industry-adjusted. ROA -89.406*** (0.000) -92.710*** (0.000) -51.907** (0.026) -53.511** (0.023)
Log(Number of analyst following) -5.924* (0.061) -5.831* (0.067) -10.790 (0.143) -10.513 (0.159)
Institutional holdings -9.483 (0.226) -10.184 (0.199) -3.576 (0.828) -11.146 (0.496)
High-yield spread 0.101*** (0.005) 0.096*** (0.008) 0.161* (0.084) 0.145 (0.123)
Number of observations 3,826 3,826 855 855
Adjusted R2 0.544 0.537 0.564 0.553
39
Table IX. Do the equity holdings by non-bank syndicate members affect the pricing of the loan facility?
This table presents the OLS regression coefficient estimates of Equation (1) and corresponding p-values with
indicator variables denoting whether the type of non-bank institution also owned at least 0.1% of the firm’s
outstanding equity during the one-year prior to the origination of the loan (non-bank syndicate member is a dual-
holder). Column (1) includes an indicator variable measuring whether any type of non-bank syndicate member is a
dual-holder. Column (2) includes indicator variables denoting the type of non-bank institution and whether that type
of non-bank institution is a dual-holder. Column (3) includes indicator variables denoting the non-bank syndicate
member is the largest lender and a dual-holder. Definitions of all variables are provided in Appendix A. The
dependent variable is the all-in-drawn loan spread over LIBOR in basis points, and the analysis is conducted at the
loan facility level. All specifications include facility-type fixed effects, facility-purpose fixed effects, and year fixed
effects. Standard errors are clustered at the firm level. ***, **, * correspond to statistical significance at the 1%, 5%,
and 10% level, respectively.
Facilities with non-bank syndicate members
Dependent Var.= All-in-drawn spread (1) (2) (3)
coef. (p-val) coef. (p-val) coef. (p-val)
Non-bank syndicate member is a dual-holder 9.159 (0.178)
32.710* (0.078)
Insurance company syndicate member
-0.378 (0.957)
Insurance company syndicate member is a dual-holder -9.315 (0.676)
Finance company syndicate member
4.162 (0.483)
Finance company syndicate member is a dual-holder -7.836 (0.403)
HF/PE syndicate member
29.288*** (0.000)
HF/PE syndicate member is a dual-holder 28.460 (0.120)
MF syndicate member
7.671 (0.253)
MF syndicate member is a dual-holder -1.939 (0.833)
Other non-bank syndicate member
29.244*** (0.000)
Other non-bank syndicate member is a dual-holder
-49.216* (0.059)
Non-bank syndicate member is the largest lender
22.911* (0.058)
Non-bank syndicate member is the largest lender and
a dual-holder -28.837 (0.287)
Log(Facility amount) -26.771*** (0.000) -26.072*** (0.000) -12.351** (0.025)
Log(Number of participating lenders) -15.266*** (0.000) -21.709*** (0.000) -11.709 (0.191)
% of participating lenders with past relationship -25.320*** (0.000) -15.612** (0.029) -19.516 (0.176)
Log(Maturity) -20.602*** (0.000) -19.799*** (0.000) -29.327*** (0.007)
Secured Indicator 20.314** (0.010) 18.152** (0.020) 38.862*** (0.004)
Second-lien Indicator 308.408*** (0.000) 302.906*** (0.000) 307.149*** (0.000)
Covenants Indicator -1.214 (0.883) 0.601 (0.941) -29.019 (0.142)
Syndicated facility Indicator -16.347 (0.246) -16.225 (0.251) -41.372* (0.066)
Log(Total assets) -1.282 (0.637) -1.430 (0.594) -10.413* (0.087)
Fixed assets/total assets 21.216* (0.061) 18.271* (0.098) 3.716 (0.860)
Z-score -3.290*** (0.003) -3.009*** (0.007) -5.522*** (0.006)
Leverage 11.699 (0.180) 10.761 (0.216) 17.794 (0.329)
Industry-adjusted ROA -69.581*** (0.000) -66.632*** (0.000) -54.158** (0.021)
Log(Number of analyst following) -8.142** (0.044) -6.848* (0.082) -11.813 (0.115)
Institutional holdings -15.271* (0.088) -12.887 (0.139) -10.894 (0.507)
High-yield spread 0.090** (0.040) 0.090** (0.040) 0.140 (0.142)
Number of observations 3,460 3,460 855
Adjusted R2 0.487 0.501 0.553
40
Table X. Uses of loan facility proceeds
This table presents the results from the following regression specification:
εYearQtrθassetstotallnβresidualSpreadβ
residualSpread1assetstotal
proceedloanlnβ1
assetstotal
proceedloanlnββ1
asetstotal
cashcashln
20074Q
19971Qi
i043
0
2
0
10
0
0t
t=1, 4, 8 corresponds to the fiscal quarter following the issuing quarter. All regressions include year-quarter fixed effects. Loan facility proceeds are aggregated
within a calendar-quarter. Dollar changes are the implied change in the dependent variable when loan facility proceeds are increased by $1 (calculations are based
on a median-sized firm/facility in the sample. Year-quarter fixed-effects are for 2003Q3). ***, **, * correspond to statistical significance at the 1%, 5%, and 10%
level, respectively.
1 2 3 4 $Change
t N Coeff. (p-val) Coeff. (p-val) Coeff. (p-val) Coeff. (p-val)
Median
Spread
Residual
Median
Spread
Residual + σ
adj-R2
Wald-test
(H0: 2,Non-bank
= 2, Bank-only)
Panel A: All facilities
1Q 7,816 0.1025 (0.208) -0.0004 (0.321) 0.0001 (0.290) 0.0044 (0.069) 0.088 0.062 0.032
4Q 7,338 0.1154 (0.146) -0.0007* (0.091) 0.0002* (0.070) 0.0027 (0.287)
0.101 0.050 0.034
8Q 6,698 0.0421 (0.337) -0.0009* (0.077) 0.0002* (0.052) -0.0050*** (0.008) 0.045 -0.021 0.018
Panel B: Non-bank facilities
1Q 2,019 0.1654 (0.208) -0.0018 (0.187) 0.0005 (0.199) 0.0020 (0.363)
0.122 -0.025 0.110 3.63* (0.057)
4Q 1,872 0.1564 (0.257) -0.0020 (0.148) 0.0005 (0.166) -0.0020 (0.552)
0.106 -0.056 0.106 3.41* (0.065)
8Q 1,686 0.1704 (0.248) -0.0019 (0.200) 0.0005 (0.194) -0.0042 (0.187)
0.116 -0.033 0.112 0.45 (0.502)
Panel C: Bank-only facilities
1Q 5,797 0.1184 (0.319) 0.0004 (0.238) -0.0001 (0.419) 0.0058* (0.096) 0.091 0.118 0.034
4Q 5,466 0.1417 (0.208) 0.0001 (0.903) 0.0000 (0.843) 0.0051 (0.149)
0.117 0.121 0.033
8Q 5,012 0.0013 (0.968) -0.0008 (0.209) 0.0002 (0.103) 0.0058** (0.043) 0.017 -0.038 0.010
41
Appendix A: Variable definitions
Variables Definition
Non-bank syndicate member An indicator variable that takes a value of one if the loan facility least one
non-bank (neither commercial bank nor investment bank) institutional
syndicate member, and zero otherwise.
Source: DealScan
Non-bank syndicate members’ loan share Sum of loan shares held by non-bank (neither commercial bank nor
investment bank) institutional investors.
Source: DealScan
Non-bank syndicate member is the largest lender An indicator variable that takes a value of one if non-bank (neither
commercial bank nor investment bank) institutional investor(s) funded the
largest share of the facility, zero otherwise.
Source: DealScan
Non-bank syndicate member is a dual-holder An indicator variable that takes a value of one if the loan facility has the
participation of at least one non-bank (neither commercial bank nor
investment bank) syndicate member who held at least 0.1% of equity stake
in the same borrowing company during the1-year period leading up to the
current loan, and zero otherwise.
Source: DealScan, Thompson Reuters Institutional Holding
Non-bank syndicate member is the largest lender
and a dual-holder
An indicator variable that takes a value of one if non-bank (neither
commercial bank nor investment bank) institutional syndicate member(s)
in the facility is the largest lender and also held at least 0.1% of equity
stake in the same borrowing company during the1-year period leading up
to the current loan, and zero otherwise.
Source: DealScan, Thompson Reuters Institutional Holdings
All-in-drawn spread Basis point spread over LIBOR plus the annual fee and the up-front fee
spread, if there is any-.
Source: DealScan
Log(Facility Amount) Natural log of the facility size.
Source: DealScan
Log(Number of participating lenders ) Natural log of the number of participating lenders in the facility syndicate.
Source: DealScan
% of participating lenders with past relationship The portion of lenders in the loan facility syndicate that have made loans to
the borrower within the 36-month period prior to the current loan.
Source: DealScan
Log(Maturity) Natural log of the maturity of the facility in months.
Source: DealScan
Secured Indicator An indicator variable that takes a value of one if the facility is secured, and
zero otherwise.
Source: DealScan
Second-lien Indicator An indicator variable that takes a value of one if the facility is second-lien,
and zero otherwise.
Source: DealScan
Covenant Indicator An indicator variable that takes a value of one if the loan has covenants,
and zero otherwise.
Source: DealScan
Syndicated Indicator An indicator variable that takes a value of one if the loan is distributed to a
syndicate of lenders, and zero otherwise.
Source: DealScan
(Continues to the next page)
42
(Appendix A continued from the previous page)
Revolver Indicator An indicator variable that takes a value of one if the facility type is
revolving line of credit (Revolver/Line, Revolver, 364-Day Facility,
Demand Loan, Limited Line in DealScan), and zero otherwise.
Source: DealScan
Term Loan A facility Indicator An indicator variable that takes a value of one if the facility type is Term
Loan A facility, and zero otherwise.
Source: DealScan
Term Loan B facility Indicator An indicator variable that takes a value of one if the facility type is Term
Loan B facility or higher (C, D, …., H) or unlabeled, and zero otherwise.
Source: DealScan
Log(total assets) Natural log of the total assets of the borrower at the end of fiscal year prior
to the current loan.
Source: Compustat
Fixed assets/Total assets The borrower's asset tangibility at the end of fiscal year prior to the current
loan, calculated as Net Property, Plant, and Equipment (PP&E)/total assets
Source: Compustat
Z-score Altman's Z-score for the borrower at the end of fiscal year prior to the
current loan.. Z-score is calculated as Z=1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 +
0.99X5, where X1 is working capital/total assets, X2 is retained
earnings/total assets, X3 is EBIT/total assets, X4 is market value of
equity/book value of total liabilities, and X5 is sales/total assets (Altman
(1968)).
Source: Compustat
Leverage The borrower's book leverage at the end of fiscal year prior to the current
loan, calculated as book value of total debt/book value of total assets.
Source: Compustat
Industry-adusted ROA The borrower's ROA in excess of the median of the corresponding 2-digit
SIC industry ROA at the end of fiscal year prior to the current loan.
Source: Compustat
Log(Number of analyst following) Natural log of the number of analysts following the borrower's stock.
Missing values are coded as zero.
Source: I/B/E/S
Institutional holdings The sum of the borrower's stock held by all institutional investors at the
end of fiscal year prior to the current loan. Missing values are coded as
zero.
Source: Thompson Reuters Institutional Holdings
S&P Issuer Rating The borrower’s S&P long-term domestic issuer credit rating. A lower
value corresponds to a lower rating, with the highest rating (AAA)
receiving a value of 22 and the lowest rating (D) receiving a value of 1.
Missing ratings are assigned a value of zero.
Source: Compustat
High-yield Spread Market credit spread in the month of loan issuance. The credit spread is
measured as (Bank of America Merrill Lynch US Corporate High Yield
BB Option-Adjusted Spread – Bank of America Merrill Lynch US
Corporate AAA Option-Adjusted Spread) in basis points.
Source: Federal Reserve Bank
43
Appendix B: Summary statistics
This table presents summary statistics for the final sample that has a full set of data (12,346 facilities). All
continuous variables are winsorized at the 1% and 99% level.
N Mean 25th Pct. Median 75th Pct. Std. Dev.
Panel A: Facility characteristics
Facility amount ($MM) 12,346 159 17 65 200 253
Number of participating lenders 12,346 6.18 1.00 3.00 8.00 7.31
% of participating lender with past relationship 12,346 30.53 0.00 9.52 57.14 37.08
Maturity (months) 12,346 45.68 27.00 47.00 60.00 23.12
Secured indicator 12,346 0.731 0.000 1.000 1.000 0.443
Second-lien indicator 12,346 0.010 0.000 0.000 0.000 0.099
Covenant indicator 12,346 0.821 1.000 1.000 1.000 0.383
Syndicated indicator 12,346 0.877 1.000 1.000 1.000 0.329
Revolver indicator 12,346 0.653 0.000 1.000 1.000 0.476
All-in-drawn spread 12,346 249.4 155.0 225.0 300.0 127.4
Non-bank participated 12,346 0.280 0.000 0.000 1.000 0.449
Panel B: Borrowing firm characteristics
Total Assets ($MM) 12,346 1,640 96 333 1,128 4,053
Fixed assets/total assets 12,346 0.314 0.120 0.251 0.462 0.236
Z-score 12,346 3.170 1.315 2.462 3.992 3.684
Leverage 12,346 0.598 0.393 0.562 0.731 0.329
Industry adjusted ROA 12,346 -0.080 -0.107 -0.030 0.015 0.202
Number of analyst following 12,346 4.415 1.000 2.833 6.333 4.884
Institutional holdings 12,346 0.427 0.124 0.407 0.702 0.321
Has S&P issuer credit rating 12,346 0.426 0.000 0.000 1.000 0.494
S&P issuer credit rating (conditional on having a credit rating) 5,255 9.944 9.000 10.000 11.000 2.658
Panel C: % of average loan share (conditional on participation)
All bank syndicate members 3,664 94.8 100.0 100.0 100.0 14.3
Commercial bank 3,620 90.0 85.5 100.0 100.0 17.7
Investment bank 1,034 21.0 9.1 15.1 25.0 20.0
All non-bank syndicate members 855 44.2 13.3 33.3 80.0 35.8
Insurance company 77 11.2 4.0 8.0 13.3 13.6
Finance company 654 35.3 8.2 20.0 50.0 34.7
HF/PE 297 34.5 9.2 19.6 50.0 33.9
Mutual fund 111 16.6 4.0 9.5 20.0 19.0
Other non-bank syndicate members 78 22.4 2.5 6.5 25.8 32.2
Panel D: % of being largest lender (conditional on participation)
All bank syndicate members 3,664 96.6 100.0 100.0 100.0 18.2
Commercial bank 3,620 94.3 100.0 100.0 100.0 23.3
Investment bank 1,034 28.0 0.0 0.0 100.0 44.9
All non-bank syndicate members 855 47.0 0.0 0.0 100.0 49.9
Insurance company 77 7.8 0.0 0.0 0.0 27.0
Finance company 654 41.4 0.0 0.0 100.0 49.3
HF/PE 297 34.3 0.0 0.0 100.0 47.6
Mutual fund 111 19.8 0.0 0.0 0.0 40.0
Other non-bank syndicate members 78 20.5 0.0 0.0 0.0 40.6
44
Appendix C: Spread and payment schedule by facility type
Facility type
Revolver Term A Term B Term Loans
Total number of facilities in sample 12,421 956 2,890 3,764
Avg. all-in-drawn spread
Mean 230 271 311 301
Median 225 250 275 275
Payment Schedule N (%) N (%) N (%) N (%)
Payment Schedule information available 27 (0.2) 558 (58.4) 1,655 (57.3) 1,750 (46.5)
Payment Period
Bullet / Final Payment 3 (11.1) 61 (10.9) 707 (42.7) 397 (22.7)
Annually 2 (7.4) 9 (1.6) 11 (0.7) 33 (1.9)
Semi-annually 0 (0.0) 25 (4.5) 59 (3.6) 49 (2.8)
Quarterly 16 (59.3) 432 (77.4) 815 (49.2) 798 (45.6)
Monthly 6 (22.2) 29 (5.2) 59 (3.6) 463 (26.5)
Other 0 (0.0) 2 (0.4) 4 (0.2) 10 (0.6)
Figure A.1. Distribution of Payment Period by Facility Type
(11.1) (10.9)
(42.7)
(22.7) (7.4)
(1.6)
(0.7)
(1.9)
(4.5)
(3.6)
(2.8)
(59.3)
(77.4)
(49.2)
(45.6)
(22.2)
(5.2) (3.6)
(26.5)
(0.4) (0.2) (0.6)
0%
20%
40%
60%
80%
100%
Revolver Term A Term B Term Loans
Bullet/Final Pmt Annually Semi-annually Quarterly Monthly Other